Conventional methods for soil sampling and analysis for soil variability in chemical characteristics are too time-consuming and expensive for multi-seasonal monitoring over large-scale areas.
In many instances it is desirable to be able to detect the presence of nutrients in soil, particularly phosphorus, which can enrich nearby lakes in phosphorus via surface water run-off, thereby promoting cyanobacteria blooms in lakes. Showing where soils are less enriched in nutrients also can be used in precision farming to direct the farmer to add only the necessary amount of nutrients when fertilizing soils, thereby reducing the cost of fertilizer to the farmer and reducing the nutrient load added to nearby lakes from surface water run-off from farm fields.
It is particularly desirable to be able to detect the presence of nutrients in soil in a manner that is convenient and provides relatively immediate results so that the public, farmers or agricultural authorities may be warned or other actions taken to avoid or eliminate contamination of the assayed soil.
Application of treated sewage sludges (biosolids) to agricultural land has become a prominent and acceptable method of waste disposal in recent years. Biosolids are known to improve soil physical characteristics (Epstein et al., 1975; Wei et al., 1985), increase the organic matter and cation exchange capacity and supply the nutrients required for crop growth (Sommers, 1977; Singh and Agrawal, 2008). However, the potential for excess application of biosolids, resulting in a build up of nitrogen, phosphorus (Mantovi et al., 2005), zinc, copper, lead (Mantovi et al., 2005; Udom et al., 2004; Nyamangara and Mzezewa, 1999) and cadmium (Bergkvist et al., 2003) in the surface soils of agricultural fields continues to be an area of concern. Other types of fertilizers can build up the nutrients in soils to an impractical, potentially harmful level. Accumulation of phosphorus at high concentrations is a major environmental concern, as it affects the water quality of lakes and rivers in the event of runoff (Shober and Sims, 2003).
Hence, there is an increasing need to continuously monitor the extent of soil contamination in biosolid-applied fields, and in other types of fertilized fields, also. Even though conventional methods of soil sampling and testing are being used for this purpose, they are often expensive, time-consuming and unsuitable for mapping soil contamination over large areas.
Remote sensing has been used as an alternative method for determining and mapping the physical and chemical characteristics of the soil. High resolution aerial imagery was used to map the organic carbon (Chen et al., 2000), clay content (Sullivan et al., 2005), organic matter and Bray-1 phosphorus concentration (Varvel et al., 1999) in bare soils. Dematte et al. (2003) reported that chemical variations in soil resulting from fertilizer applications can be detected, based on the intensity of reflectance. Several studies showed the use of spectral reflectance to determine the soil color (Post et al., 2000), texture and particle size distribution (Chang et al., 2001), soil moisture (Lobell and Asner, 2002), iron oxides (Ji et al., 2002), carbonates (Ben-Dor and Banin, 1990), clay (Ben-Dor and Banin, 1995), organic carbon (Dalal and Henry, 1986; Morra et al., 1991; Reeves et al., 2002) organic matter (Henderson et al., 1992) and soil phosphorus (Bogrekci and Lee, 2005, 2007).
As used herein, remote sensing refers to the capability of obtaining information about an object without touching it. Sensors which are not in direct contact with the object are generally used to obtain the information. In a more limited context, the information obtained by remote sensing is a function of energy emitted by, absorbed by, or reflected from the object.
As used herein, where the remote sensing is conducted from distances from which vegetation may cover portions of the soil surface, vegetation may be excluded according to a masking formula, such as elimination from consideration of any areas that exceed a lower threshold of the Normalized Difference Vegetation Index (NDVI) standard, which is a simple numerical indicator that can be used to analyze remote sensing measurements, typically but not necessarily from a space platform, and assess whether the target being observed contains live green vegetation or not. This may be arrived at by using LANDSAT TM bands 3 and 4 in a formula (4−3)/(4+3) to arrive at a value that is broadly between 0.1 and 0.3, preferably about 0.2. Regions above this threshold range would be blacked out or otherwise eliminated from the algorithmic calculations herein to avoid erroneous results. Regions lower than this threshold range may be considered to be sufficiently bare soils from which accurate measurements and calculated results may be taken.
The addition of soil contaminants as a result of biosolid application tends to be concentrated in surface soil samples (Mantovi et al., 2005; Bergkvist et al., 2003; Udom et al., 2004; Nyamangara and Mzezewa, 1999).
However, there remains a need for improved methods of remote determination of soil nutrients that offer reduced expense, time savings and availability of mapping soil contamination over large areas.